Tetracycline-Regulated Transactivators Driven by the InvolucrinPromoter to Achieve Epidermal Conditional Gene Expression

Jean Jaubert, Satyakam Patel, Jun Cheng, and Julia A. SegreNational Human Genome Research Institute, NIH, Bethesda, Maryland, USA

To achieve conditional gene expression in the differentiated layers of the epidermis, we generated transgenic

mice with the tetracycline-regulated transactivator proteins, tTA (tetracycline transactivator) and rtTA (reverse

tetracycline transactivator), expressed from the human involucrin promoter. Interaction with tetracycline turns off

or turns on the tTA and rtTA molecules, respectively, allowing for regulation of downstream target genes during

development and postnatally. These transactivator lines were crossed with reporter mice driving LacZ expression

from a tetracycline response element to analyze the specificity and levels of target gene expression. Quantitative

b-galactosidase experiments demonstrate a 30-fold induction, specific to epithelial tissues. Immunohistochemistry

results illustrate that the b-galactosidase staining follows that of endogenous involucrin expression. Induction

initiates at embryonic day 14.5 with expression over the entire epidermal surface by E16.5. Together with other

driver lines, expressing tetracycline transactivators in the mitotically active layers of the epidermis, these mice

will allow investigators to specifically modulate expression of target genes to specific stages of epidermal

differentiation.

Key words: conditional gene expression/differentiation/epidermis/involucrin promoter/tetracycline-regulatedJ Invest Dermatol 123:313 –318, 2004

Toward understanding the in vivo function of proteins, onepowerful approach is to investigate the effect of turning agene on or off in temporally and spatially regulated fashion.Conditional expression is achieved with a tissue-specificpromoter driving an effecter molecule, regulated by tetracy-cline, to activate a target transgene. This system exists intwo formats: in the tet OFF system, the tetracycline trans-activator (tTA) binds DNA in the absence of tetracycline,whereas in the tet ON system, the reverse tTA (rtTA) bindsDNA in the presence of tetracycline. Regulating a gene froma tet response element (TRE), multiple levels of ectopic exp-ression can be investigated by crossing with mice that expresstetracycline transactivators in distinct layers of epidermis. Thisbi-transgenic system also circumvents the inability to estab-lish lines due to embryonic or perinatal lethality. (Reviewedin Lewandoski, 2001; Corbel and Rossi, 2002.).

Mammalian epidermis, a stratified epithelium, is a self-renewing tissue composed of a population of mitoticallyactive cells in the innermost basal layer and their derivativesthat travel upward to the skin surface in a linear program ofterminal differentiation (Fuchs and Raghavan, 2002; Nie-mann and Watt, 2002). Mice with promoters specific to thebasal layer of the epidermis expressing both tet ON andOFF have been characterized (Keratin 14 (K14): (Xie et al,1999), Keratin 5 (K5): (Diamond et al, 2000)). There are,however, no mice with tetracycline effecter moleculesexpressed from suprabasal promoters.

Involucrin is a marker of keratinocyte terminal differentia-tion and is a protein precursor of the terminally differentiatedcornified envelope (CE). Involucrin was one of the firstprecursors of the human CE to be identified and molecularlycloned (Banks-Schlegel and Green, 1981; Eckert and Green,1986). It is a soluble cytoplasmic protein, rich in glutamineand glutamic acid, that is a substrate for transglutaminase.Human involucrin is expressed in the suprabasal cells of theepidermis and other stratified epithelia, including bladderand tongue (Rice and Green, 1979). Although the sites ofinvolucrin expression in mouse are nearly identical, theonset of expression is slightly later in the upper spinouslayers and inner root sheath of the hair follicle (Djian et al,1993; Li et al, 2000). Similar to human, mouse involucrin isdetected in tissues with stratified epithelia, includingtongue, palate, uterus, esophagus, and bladder. High levelsof expression are also detected in hyperproliferative epi-dermis during wound healing (Li et al, 2000).

Regulation of the human involucrin promoter has beenextensively studied both in vitro and in vivo (Carroll andTaichman, 1992; Carroll et al, 1993; Crish et al, 1998, 2002).Maximal expression in vitro required 3.7 kb of sequence,including the first intron (1188 bp), first non-coding exonand transcription start site (70 bp) and 2473 bp of upstreamsequence (Carroll and Taichman, 1992). Appropriate epithe-lial expression is observed when this sequence drives exp-ression of the b-galactosidase enzyme in transgenic mice(Carroll et al, 1993). Further delineation demonstrated thatthe distal region (�2473 to �1953) is necessary for surfaceepithelial expression and the proximal region (�986 to �41)is necessary for internal epithelial expression (Crish et al,

Abbreviations: Dox, doxycycline; DT, double transgenics; INVp,involucrin promoter; rtTA, reverse tetracycline transactivator; TRE,tet response element; tTA, tetracycline transactivator

Copyright r 2004 by The Society for Investigative Dermatology, Inc.

313

1998). Analysis from the Eckert laboratory has identifiedmany of the binding sites and proteins that regulate theexpression of the human involucrin promoter (Crish et al,2002 and references within).

To produce conditional expression of genes in the supra-basal layer of the epidermis, we generated transgenic mouselines in which the human involucrin promoter (INVp) drivesexpression of the tetracycline-regulated transactivators, tTA,and rtTA. To characterize the levels of induction of target geneactivation as well as the spatial and temporal expression, wecrossed these lines to reporter mice with the LacZ generegulated by a TRE (TRE-LacZ) (Furth et al, 1994).

Here we show that INVp-tTA/TRE-LacZ and INVp-rtTA/TRE-LacZ double transgenic mice induced by the absenceor presence of tetracycline, respectively, express b-galacto-sidase specifically in the spinous and granular layers of theepidermis. This expression pattern is similar to the resultspreviously obtained with INVp-LacZ transgenic mouse andconfirms the specificity of the tetracycline double-trans-genic system (Carroll et al, 1993). We extend these studiesby investigating the expression of the involucrin promoterduring development. These mice are publicly available andin conjunction with other tetracycline-inducible driver miceshould allow the community to address the role of proteinsin specific layers of the epidermis.

Results

Generation and initial characterization of INVp-rtTA andINVp-tTA mice To create mice expressing the tetracyclinetransactivator (tTA) and reverse tTA (rtTA) in the suprabasallayers of the epidermis, these genes were cloned down-stream of the human involucrin promoter. This promotercontains the distal and proximal promoter regions, the firstnon-coding exon and the first intron of the human involucringene. This promoter also contains approximately 150 bp ofnon-coding SV40 sequence including a small intron (Carrollet al, 1993). tTA is a fusion of the tet repressor protein (tetR)with the VP16 activation domain (AD) to activate TRE in theabsence of tetracycline. The re-engineered rtTA protein is afusion of a mutated tetR with a trimerized minimal VP16 ADthat activates TRE in the presence of tetracycline (Fig 1A). There-engineered rtTA molecule has increased protein stabilityand sensitivity to tetracycline with reduced basal activity.

We first determined which lines of PCR positive miceexpressed the effecter mRNA and selected the two INVp-tTAand INVp-rtTA lines with highest expression for completeanalysis (Fig 1B). The tTA (1.5 kb) and rtTA (1.2 kb) transcriptsare different sizes because of their activation domains. TheINVp-tTA lines express higher levels of transactivator mRNAas compared to the INVp-rtTA. This may reflect, however, anascertainment bias since 23 and 5 founders were identifiedfor the INVp-tTA and INVp-rtTA lines, respectively.

Transactivation by INVp-rtTA and INVp-tTA is tissuespecific and regulated by doxycycline (Dox) To analyzethe specificity and levels of target gene expression, INVp-tTA and INVp-rtTA mice were crossed with reporter miceexpressing the LacZ gene from a TRE (TRE-LacZ) (Furthet al, 1994). To study relative levels of activation of a targetgene, newborn skin from induced and uninduced INVp-tTA/

TRE-LacZ (INVp-tTA-Z) double transgenics (DT) and INVp-rtTA/TRE-LacZ (INVp-rtTA-Z) DT were analyzed for expres-sion of the LacZ mRNA. When induced, the expression ofLacZ mRNA relative to G3PDH as control for loading is asfollows: INVp-rtTA line INV3: 1.00; INVp-rtTA line INV4: 1.50;INVp-tTA line I3:3.83; INVp-tTA line I4: 3.63. Note that theINVp-rtTA tet-ON lines are denoted as INV3 and INV4 andthe INVp-tTA tet-OFF lines are denoted as I3 and I4. Com-paring the highest expressors for each construct, INVp-tTA line I3 express approximately 2.5 fold higher levels ofLacZ mRNA than the INVp-rtTA line INV4 (Fig 2A). Underuninduced conditions, none of the lines activates appreciableamounts of LacZ transcription (data not shown). The levels ofLacZ activity are proportional but not strictly correlated to thelevels of transactivator transcript seen in Fig 1B.

To analyze tissue specificity and regulation by Dox oftarget gene activation, b-galactosidase enzyme activity isassayed in involucrin expressing (skin, tongue) and non-expressing tissue (liver) of INVp-tTA-Z and the INVp-rtTA-ZDT in the presence and absence of Dox. Skin samplesfor every line show at least 30-fold induction of theb-galactosidase enzyme in the induced versus uninducedstate. The levels of expression in the uninduced state wereessentially at background levels, suggesting extremely tightcontrol of induction. As was observed for the levels of LacZmRNA, the highest levels of expression were from INVp-rtTA-Z DT line INV4 and INVp-tTA-Z DT line I3. Similarly, theINVp-tTA-Z DT line I3 skin sample had approximately 2-foldhigher levels of b-galactosidase enzyme activity than theINVp-rtTA-Z DT line INV4 (Fig 2B). Utilizing different pro-moters, previous comparisons between tet-OFF and tet-ONmice have noted an almost 10-fold difference in the levels ofexpression. One possible explanation for the higher levels of

Figure1Generation and characterization of INV-tTA and INV-rtTA trans-genic mice. (A) Diagram of the INVp-tTA and INVp-rtTA transgeneconstructs. The 3.9 kb involucrin promoter was cloned upstream of thetTA and rtTA effecter molecules. Size of each part of the construct isgiven below the construct. TetR: Tetracycline repressor; FFF AD:trimerized minimal VP16 activation domain; VP16 AD: VP16 activationdomain. SBI: SnaBI, ERV: EcoRV, HIII: HindIII, SI: SalI, SII: SacII, BZI:BstZ17I, RI: EcoRI. (B) Northern blot analysis of tTA and rtTA expressionlevels for INV3 and INV4 (rtTA at 1.2 kb) and I3 and I4 (tTA at 1.5 kb)hybridized with a probe for SV40 poly A. G3PDH hybridization is usedto normalize the levels of mRNA for each lane.

314 JAUBERT ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

b-galactosidase enzyme activity for induced INVp-rtTA-ZDT mice may be the reengineered rtTA molecule.

Transactivation of the LacZ promoter also occurred in thetongue, which expresses involucrin, but to almost a 10-foldlesser extent than in the skin (Fig 2B). This may reflect thelower percent of Involucrin positive cells per gram of tissue inthe tongue than in the skin. For all of the samples, the level ofb-galactosidase enzyme in the liver, a non-involucrin-expres-sing tissue, was similar to background levels (data not shown).

Under identical conditions, we directly compared the le-vels of luminescence in the skin of INVp lines with the Glicklaboratory K5p-tTA line 1216, which yielded the highestlevels of b-galactosidase enzyme of all lines reported. The

K5p-tTA produced 5.9-fold higher levels of activation thanthe INVp-tTA line I3 (821 vs 138 RLU per mg protein) and11.6-fold higher levels of activation than the INVp-rtTA lineINV4 (821 vs 70). The variance between these lines mayreflect the number of cells expressing the promoter or thestrength of the promoter. Importantly, these driver micehave levels of activation within an order of magnitude ofeach other, allowing an investigator to modulate the levelsand patterns of expression of specific proteins.

Dose dependence of INVp-rtTA-Z and INVp-tTA-Z DT Toanalyze the levels of Dox necessary to achieve activation ofthe INVp-rtTA-Z DT (tet-ON), INV4 mice were provided witha 25-fold range dose of Dox for 1 wk; i.e., 0.4, 2, or 10 g perliter. At the end of this week, the levels of LacZ expressionwere quantified from a skin biopsy. For all three doses, thelevel of LacZ expression was similar and statisticallyindistinguishable (46 RLUper mg protein). This is consistentwith the prediction from the original report by Bujard’slaboratory that the newly engineered rtTA-M2 molecule willbe induced to the same level with 10-fold lower Dox thanwas previously utilized. These INVp-rtTA-Z DT mice wereswitched to regular drinking water and biopsies were takenat day 2 and day 5. Only the animals kept on the lowestlevels (0.4 g per liter) showed a significant decline in thelevel of expression of LacZ by day 2 (12 RLU per mg protein).Mice on the medium and high levels showed a modestdecline by day 2 (28 RLU per mg protein and 36 RLU per mgprotein for 2 and 10 g per liter, respectively). Finally, by day 5mice receiving all three doses had significant declines inLacZ gene expression almost reaching background levelsof expression (7 RLU per mg protein). These results indicatethat all three levels of Dox (0.4, 2, and 10 g per liter) will achi-eve the same level of activation. If an investigator would liketo reverse the downstream gene activation, however, thiscan most rapidly be achieved with the lowest level of Dox.

This same experiment was performed on the INVp-tTA-ZDTwith 1 w on the 25-fold range of Dox (0.4, 2, and 10 g perliter) and switched to regular water while monitoring geneexpression from biopsied skin. Again, all three levels of Doxwere able to achieve complete suppression of gene ex-pression, indicating that the lowest level of Dox is sufficientto regulate downstream gene expression. For the INVp-tTA-Z DTs, however, there was still no gene expression 5 d afterthe switch to regular drinking water. This continuingrepression of the INVp-tTA-Z DT even with low levels ofDox may prove difficult for investigators who want to rapidlyturn back on gene expression.

Localization of b-galactosidase expression in INVp-rtTA-Z and INVp-tTA-Z DT mice mimics endogenous involu-crin expression Consistent with the staining data of adultmice from the Watt laboratory, the involucrin antibodyreacts with protein in the granular layer of the epidermis andthe inner root sheath of the hair follicle of the newbornmouse (Fig 3A) (Li et al, 2000). When crossed with TRE-LacZ mice and induced, both newborn INVp-tTA-Z andINVp-rtTA-Z DT show b-galactosidase enzyme activity con-sistent with the expression of the endogenous involucrinprotein (Fig 3B, C). There are distinctions, however, betweenthe DT-induced mice. First, the INVp-tTA-Z DT have positive

Figure 2Differences in the levels of induction obtained between lines ofINVp-rtTA-Z and INVp-tTA-Z DT mice. Levels of transactivation of theINVp-rtTA and INVp-tTA mice are ascertained by crossing them with theTRE-LacZ responder mice with or without Dox. (A) Northern blotanalysis of LacZ mRNA (4.2 kb) expression from skin of inducednewborn INVp-rtTA-Z DT (lines INV3 and INV4) and INVp-tTA-Z DT(lines I3 and I4). G3PDH hybridization is used to normalize the levels ofmRNA for each lane. (B) Tissue-specific transactivation of b-galacto-sidase enzyme is activated by Dox in INVp-rtTA-Z DT (lines: INV3 andINV4). On the reverse, tissue-specific transactivation of b-galactosi-dase enzyme is induced by the absence Dox in INVp-tTA-Z DT (lines: I3and I4). Tissue type, line, and presence or absence of Dox is indicatedon the X-axis. The shaded bars are the readings in the induced stateand the unshaded bars are the uninduced state. The height of the barrepresents the average of four mice with standard deviations indicated.

CONDITIONAL GENE EXPRESSION IN DIFFERENTIATED EPIDERMIS 315123 : 2 AUGUST 2004

cells in the lower spinous cells, whereas the INVp-rtTA-Z DTexpression is constrained to the upper spinous cells. Sec-ond, only the INVp-tTA-Z DTs show prominent expression inthe hair follicles. Third, the time required for strong stainingof the sections is less for the INVp-tTA lines than for theINVp-rtTA, consistent with the levels of activation observedabove. When the INVp-rtTA-Z and INVp-tTA-Z DTs areuninduced, no b-galactosidase enzyme activity is detectedin situ (data not shown). The expression in the tongue ofINVp-rtTA-Z and INVp-tTA-Z DTs is also faithful to endo-genous involucrin, with staining in suprabasal cells of boththe dorsal and ventral surfaces of the tongue (Fig 3D, E).

Endogenous involucrin expression and b-galactosidaseexpression in INVp-rtTA-Z and INVp-tTA-Z DTs duringdevelopment Before analyzing the pattern of these lines indevelopment, we first investigated the endogenous involu-crin expression. We first detect involucrin mRNA at E15.5with upregulation at E16.5 (Fig 4A). Involucrin is expres-sed later than keratin 1 or 10 and after initial strati-fication but before barrier acquisition (Byrne et al, 1994;Hardman et al, 1998). Involucrin protein is expressed in the

upper spinous cells throughout development at E15.5,E16.5 and E17.5 (Fig 4B).

To analyze the pattern of expression, we performedwhole-mount staining on INVp-tTA-Z and INVp-rtTA-Z DT.Expression of the b-galactosidase enzyme is first detectedin the head and whisker pads at E14.5. The expression ofendogenous involucrin was not detected on the northernblot at E14.5 because the embryo heads were removedprior to processing the samples for mRNA. The b-gala-ctosidase enzyme signal spreads across the lateral andventral body surfaces with a striped pattern at E15.5. ByE16.5 the entire surface of the embryo possesses b-galactosidase enzyme activity (Fig 4C, D). The stripedexpression pattern at E15.5 could indicate some positioneffect variegation. The same pattern, however, was ob-served with multiple lines and from both constructs.Although the involucrin promoter driving the tTA and rtTAeffecters gave similar expression patterns, the time requiredfor staining was less for tTA than for the rtTA animals,consistent with the quantitative data presented in Fig 2.

Discussion

The bitransgenic tetracycline system enables investigatorsto query the function of proteins by ectopically expressing

Figure 3Location of Involucrin protein expression in newborn skin in wild-type mice is consistent with b-galactosidase expression in INVp-tTA-Z and INVp-rtTA-Z DT mice. (A) Anti-involucrin immunohisto-chemistry of paraffin embedded sections of wild-type newborn mouseskin. (B–E) Pattern of expression of the INV-rtTA and INV-tTA drivermice characterized by crossing with responder TRE-LacZ mice andanalyzing the b-galactosidase expression in newborn skin or tongue ofdouble transgenics, as indicated. Dotted line indicates the basementmembrane. White dotted line indicates that an area of the dermis withno signal has been cropped.

Figure4Endogenous involucrin expression during development and whole-mount b-galactosidase expression in INVp-rtTA-Z and INVp-tTA-ZDT. (A) Northern blot analysis of Involucrin mRNA expression (1.9 kb)in wild-type mice during development from E10.5 to E16.5. A signal isfirst detected at E15.5. G3PDH probe is a control for loading. (B) Anti-involucrin immunohistochemistry of paraffin-embedded sections ofwild-type E15.5, E16.5 and E17.5 mouse skin. (C, D) Developmentalexpression pattern of INVp-tTA and INVp-rtTA lines, determined bycrossing with the responder TRE-LacZ line at E14.5, E15.5, and E16.5.

316 JAUBERT ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

them in spatially and temporally regulated fashions. Thepower of this system is greatly increased when multipledriver mice are available to direct expression at discretepoints in the differentiation pathway of a specific tissue. Wereport here the characterization of tet ON and tet OFF trans-activators driven by the human involucrin promoter andexpressed in epidermal cells that have withdrawn from thecell cycle. Utilizing these mice should complement studiesperformed with keratin 5 promoter driving the tetracycline-regulated transactivators in mitotically active epidermal cells(Diamond et al, 2000). Additional lines driving expression ofthe tetracycline transactivators from the well-characterizedkeratin 1 or keratin 10 and loricrin promoters would give afull complement of expression for target genes in theepidermis (Greenhalgh et al, 1993; DiSepio et al, 1999).

By creating one line of responder mice expressing atarget gene from a tetracycline responder element, onecould then cross this line with the many driver mice for ex-pression in the epidermis. Not only would this finely dissectthe role of a given protein in the layers of the epidermis, butthis systematic study might also resolve some of thephenotypic discrepancies observed when the same proteinis expressed from different promoters. For example, twogroups investigated the ectopic expression of desmoglein 3(Dsg3) in the epidermis and found non-overlapping pheno-types (Elias et al, 2001; Merritt et al, 2002). These transgenicmice, besides having unique integration sites in the genomealso had differences in the promoter utilized (K1 vs Inv), tagon protein, background strain of mice, copy number and theutilization of mouse vs. human sequences. Creating oneresponder mouse expressing Dsg3 to mate with driver micefrom K1 and Inv promoters could standardize many of theseinherent genetic differences.

Gene activation from the tet-OFF INVp-tTA line I3 is2–2.5-fold higher than the levels achieved with the tet-ONINVp-rtTA line INV4 as measured by b-galactosidaseenzyme activity and the levels of LacZ mRNA transcription.These levels are relative to the levels of transactivatortranscript expressed by each line. These results support theresults from studies in Saccharomyces cerevisiae that thenew transactivator, rtTA-M2 results in higher levels of geneactivation with no background expression in the absence ofDox (Urlinger et al, 2000).

Although INVp-rtTA/LacZ DT showed the same level ofactivation of LacZ expression over a 25-fold range dose ofDox (0.4–10 g per liter), only the mice administered thelowest dose (0.4 g per liter) showed a significant decline inLacZ expression 2 d after the switch to regular drinkingwater. By 5 d on regular drinking water, all of the INVp-rtTA/LacZ DT expressed only background levels of LacZ.Therefore, if the investigator would like to reverse down-stream gene activation, the lowest levels of Dox (0.4 g perliter) should be used. In contrast, for the INVp-tTA-Z DT nogene expression 5 d after the switch to regular drinkingwater was observed even at the lowest levels of Dox tested.

The expression of the INVp-tTA and INVp-rtTA micemimics the endogenous expression of the involucrinpromoter with expression in the upper spinous and granularcells of the epidermis. INVp-tTA I3 gives the strongest levelof expression and is the only one to show expression in thehair follicle. Analysis of involucrin mRNA and protein

expression during development indicates that these driverlines mimic the endogenous expression.

Preliminary results from our laboratory indicate that thelevels of transactivator produced by the INVp-tTA and INVp-rtTA lines are both sufficient to activate gene expressionand produce phenotypes from our TRE-Klf4 lines. Furtherstudies to dissect the epidermal and hair follicle phenotypesof these DT are underway (J.J. and J.S. unpublished data).

Other investigators will have additional experimentalneeds for these lines and will need to determine for specificdevelopmental stages how rapidly the lines can be turnedON and OFF with the administration or removal of Dox.These results will allow investigators to specifically tailortheir experimental design and to ascertain the confidence oftheir conclusions.

These mice driving the tetracycline transactivators fromthe involucrin promoter are now publicly available to modu-late expression in skin. These mice, in conjunction withother driver lines, should prove useful to study epidermaldifferentiation and development as well as carcinogenesisand wound healing.

Materials and Methods

Generation of INVp-rtTA and INVp-tTA constructs and mice TheTet-Off protein (tTA) encodes the Escherichia coli tetracycyline tetrepressor (tetR) protein fused to the VP16 activation domain ofherpes simplex virus. The Tet-On protein (rtTA) has been re-engineered by Hillen and colleagues, who altered five amino acidsin the tetR and reduced the VP16 activation domain to a minimal 12amino acids (known as the F domain), which is trimerized to 36amino acids (FFF) (Urlinger et al, 2000). The involucrin promoterutilized contains 3.7 kb of DNA upstream of the first coding exon ofthe gene and approximately 150 bp of non-coding SV40 sequenceincluding a small intron (Carroll et al, 1993) (Fig 1A).

Human Involucrin promoter was cut from the original vector(pH3700-pL2) with SalI–NotI and cloned into the same sites inpTRE2 vector (Clontech, Palo Alto, California) to increase thenumber of available sites (Carroll et al, 1993). INVp was cut out ofthis new vector to make two constructs: (i) INVp-rtTA construct, INVpdigested with EcoRV–SacII, and subcloned in SnaBI–SacII compa-tible ends in pUHDrtTA2S-M2 (Urlinger et al, 2000); and (ii) INVp-tTAconstruct: INVp digested with EcoRV–EcoRI and subcloned inBstZ17I–EcoRI compatible ends in pTet-Off vector (Clontech). Bothconstructs were purified by CsCl gradient (Lofstrand, Gaithersburg,Maryland). To excise the insert, 30 mg of DNA was digested withHindIII, cleaving at the 50 end of the human involucrin promoter and30 to the SV40 poly A. For the INVp-tTA vector, the DNA wasadditionally digested with DrdI to make the vector backbone moreeasily distinguished from the insert for gel purification. Inserts werepurified (QIAGEN, Valencia, California) and injected into FVB/N one-cell eggs following standard pronuclear injection (Hogan et al, 1986).Positive founders were identified by PCR and maintained on FVB/Nbackground. For the INVp-tTA construct, 23 founders were identifiedas positive. For the INVp-rtTA construct, 5 founders were identifiedas positive. Genotyping of embryos was done by PCR using thefollowing primers: INVp-tTA transgene: 50-AGGGAAGAGGGGATGC-TAAA, 50-CCATCGCGATGACTTAGT; INVp-rtTA transgene: 50-AGG-GAAGAGGGGATGCTAAA, 50-AGAGCACAGCGGAATGACTT. LacZprimers were previously published (Diamond et al, 2000). All studieson animals were approved by the animal care and use committeeand made following their recommendations in our AALAC-accre-dited animal facility.

Characterization of the INV-tTA and INV-rtTA transgeneexpression INVp-tTA and INVp-rtTA lines were time-mated with

CONDITIONAL GENE EXPRESSION IN DIFFERENTIATED EPIDERMIS 317123 : 2 AUGUST 2004

mice expressing LacZ regulated by a TRE (TRE-LacZ) (gift fromL. Henninghausen and A. Glick) (Furth et al, 1994). The morningof vaginal plug detection was taken as embryonic day (E) 0.5.The mice were maintained either with or without the tetra-cycline derivative Dox as indicated in text or figures beginning atgestational day E10.5. Dox was administered in the drinking waterat 2 g per liter in 5% sucrose for all experiments except the timecourse of induction in which Dox was administered in the drinkingwater at 0.4, 2, or 10 g per liter.

RNA isolation and northern blot analyses For newborns, E15.5,and E16.5 embryos, skin from the body was isolated by dissection.At E12.5, E13.5 and E14.5, heads and inner organs were removedprior to tissue processing. At E10.5 and E11.5, heads wereremoved and bodies used for tissue sampling. All samples weresnap frozen in liquid nitrogen. Tissue was pulverized, homogenizedin Trizol (Invitrogen, Grand Island, New York) and RNA wasextracted following manufacturer’s recommendations. Approxi-mately 15 mg RNA was loaded in every lane and visualized by EtBrfor integrity of the samples. Blots were hybridized with LacZ probe(amplified by PCR on cDNA with primers: 50-ACCAGCGAAATG-GATTTTTG and 50-TAGCGGCTGATGTTGAACTG) with SV40 polyAþ probe (digested with BamHI and HindIII from rtTA2SM2 andgel purified as a 450 bp product) or with G3PDH probe (amplifiedby PCR on cDNA with primers: 50-ACCACAGTCCATGCCATCACand 50-TCCACCACCCTGTTGCTGTA) as loading control. Signalswere quantified using Molecular Dynamics Phosphorimager and IQanalysis software (Amersham, Piscataway, New Jersey).

Detection of b-galactosidase To quantitate b-galactosidase en-zyme activity, tissues were homogenized in a buffer contain-ing 100mM potassium phosphate pH7.8, 0.2% Tween 20, 1 mMdithiothreitol, and the cleared supernatant was reacted with aluminescent substrate reagent using a Luminescent b-galactosi-dase detection kit (Invitrogen). b-galactosidase activity was meas-ured using a Turner Designs 20/20 luminometer and expressed asrelative light units per microgram of protein. Readings wereperformed at 49.9% sensitivity.

Histochemical staining for b-galactosidase activity was per-formed at 371C with 5-bromo-4-chloro-3-inolyl-D-galactopyranoside(X-gal) substrate on frozen sections from newborn mice as previouslydescribed (Diamond et al, 2000). Images were acquired withOpenlab software using an Axioskop microscope (Zeiss, Thorn-wood, New York) equipped with a digital Axiocam camera (Zeiss).

Embryos were collected at E14.5, E15.5 and E16.5 and pre-fixed for 30 min in 4% paraformaldehyde. Whole-mount stainingwith X-gal substrate was performed at 371C as previouslydescribed (Byrne et al, 1994). After post-fixing in 4% paraformal-dehyde, embryos were photographed under a MZFLIII dissectingmicroscope (Leica, Bannockburn, Illinois) as above.

For the timed induction studies, 5 mm diameter skin biopsieswere removed from the dorsal surface and sutured closed.Subsequent days, biopsies were removed from an area at least2 cm from any previous area biopsied.

Immunohistochemistry Backskin samples from wild-type em-bryos were taken at E15.5, E16.5 and E17.5 and at newborn stage,fixed overnight in 4% PFA /PBS and embedded. Paraffin sectionswere hybridized with rabbit polyclonal antibodies against Involucrin(1:500) (Covance, Berkeley, California). Biotinylated secondaryantibody (1:200) was used with Vectastain ABC kit and DABsubstrate (Vector Laboratories, Burlingame, California).

The authors would like to thank Cherry Yang for assistance with theanimal care and genotyping. Michael Cichanowski assisted inpreparing the figures. Esther Fine assisted in the initial characteriza-tions of the lines. Members of the lab provided useful advice andcriticisms at all stages of the work.

DOI: 10.1111/j.0022-202X.2004.23203.x

Manuscript received August 13, 2003; revised March 10, 2004;accepted for publication March 11, 2004

Address correspondence to: Julia A. Segre, National Human GenomeResearch Institute, NIH, 49 Convent Drive, Bethesda, MD 20892, USA.Email: jsegre@nhgri.nih.gov

References

Banks-Schlegel S, Green H: Involucrin synthesis and tissue assembly by keratino-

cytes in natural and cultured human epithelia. J Cell Biol 90:732–777, 1981

Byrne C, Tainsky M, Fuchs E: Programming gene expression in developing

epidermis. Development 120:2369–2383, 1994

Carroll JM, Albers KM, Garlick JA, Harrington R, Taichman LB: Tissue- and

stratum-specific expression of the human involucrin promoter in

transgenic mice. Proc Natl Acad Sci USA 90:10270–10274, 1993

Carroll JM, Taichman LB: Characterization of the human involucrin promoter

using a transient beta-galactosidase assay. J Cell Sci 103:925–930, 1992

Corbel SY, Rossi FM: Latest developments and in vivo use of the Tet system: Ex

vivo and in vivo delivery of tetracycline-regulated genes. Curr Opin

Biotechnol 13:448–452, 2002

Crish JF, Bone F, Banks EB, Eckert RL: The human involucrin gene contains

spatially distinct regulatory elements that regulate expression during early

versus late epidermal differentiation. Oncogene 21:738–747, 2002

Crish JF, Zaim TM, Eckert RL: The distal regulatory region of the human involucrin

promoter is required for expression in epidermis. J Biol Chem 273:30460–

30465, 1998

Diamond I, Owolabi T, Marco M, Lam C, Glick A: Conditional gene expression in

the epidermis of transgenic mice using the tetracycline-regulated

transactivators tTA and rTA linked to the keratin 5 promoter. J Invest

Dermatol 115:788–794, 2000

DiSepio D, Bickenbach JR, Longley MA, Bundman DS, Rothnagel JA, Roop DR:

Characterization of loricrin regulation in vitro and in transgenic mice.

Differentiation 64:225–235, 1999

Djian P, Phillips M, Easley K, Huang E, Simon M, Rice RH, Green H: The involucrin

genes of the mouse and the rat: Study of their shared repeats. Mol Biol

Evol 10:1136–1149, 1993

Eckert RL, Green H: Structure and evolution of the human involucrin gene. Cell

46:583–589, 1986

Elias PM, Matsuyoshi N, Wu H, Lin C, Wang ZH, Brown BE, Stanley JR:

Desmoglein isoform distribution affects stratum corneum structure and

function. J Cell Biol 153:243–249, 2001

Fuchs E, Raghavan S: Getting under the skin of epidermal morphogenesis. Nat

Rev Genet 3:199–209, 2002

Furth PA, St Onge L, Boger H, et al: Temporal control of gene expression in

transgenic mice by a tetracycline-responsive promoter. Proc Natl Acad

Sci USA 91:9302–9306, 1994

Greenhalgh DA, Rothnagel JA, Wang XJ, et al: Hyperplasia, hyperkeratosis and

benign tumor production in transgenic mice by a targeted v-fos oncogene

suggest a role for fos in epidermal differentiation and neoplasia.

Oncogene 8:2145–2157, 1993

Hardman MJ, Sisi P, Banbury DN, Byrne C: Patterned acquisition of skin barrier

function during development. Development 125:1541–1552, 1998

Hogan BL, Constantini F, Lacy E: Manipulating the Mouse Embryo: A Laboratory

Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1986

Lewandoski M: Conditional control of gene expression in the mouse. Nat Rev

Genet 2:743–755, 2001

Li ER, Owens DM, Djian P, Watt FM: Expression of involucrin in normal,

hyperproliferative and neoplastic mouse keratinocytes. Exp Dermatol

9:431–438, 2000

Merritt AJ, Berika MY, Zhai W, Kirk SE, Ji B, Hardman MJ, Garrod DR: Suprabasal

desmoglein 3 expression in the epidermis of transgenic mice results

in hyperproliferation and abnormal differentiation. Mol Cell Biol 22:

5846–5858, 2002

Niemann C, Watt FM: Designer skin: Lineage commitment in postnatal epidermis.

Trends Cell Biol 12:185–192, 2002

Rice RH, Green H: Presence in human epidermal cells of a soluble protein

precursor of the cross-linked envelope: Activation of the cross-linking by

calcium ions. Cell 18:681–694, 1979

Urlinger S, Baron U, Thellmann M, Hasan MT, Bujard H, Hillen W: Exploring the

sequence space for tetracycline-dependent transcriptional activators:

Novel mutations yield expanded range and sensitivity. Proc Natl Acad Sci

USA 97:7963–7968, 2000

Xie W, Chow LT, Paterson AJ, Chin E, Kudlow JE: Conditional expression of the

ErbB2 oncogene elicits reversible hyperplasia in stratified epithelia and

up-regulation of TGFalpha expression in transgenic mice. Oncogene

18:3593–3607, 1999

318 JAUBERT ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY